of 21
7/30/2019 Xx Siecle Economics
1/21
Stavros Ioannides
Panteion UniversityDepartment of Political
Science and History
136 Syngrou Av.
176 71 Athens, GreeceEmail: [email protected]
Information and Knowledge in 20th
Century Economics:
From Prices, to Contracts, to Institutions
1st
(and very incomplete) Draft
May 2003
7/30/2019 Xx Siecle Economics
2/21
2
1 Introduction
The role of information and knowledge in economic activity has figured prominently in
most of 20th
century economics. However, economists have relied upon a very narrow
understanding of the two concepts, as illustrated by the fact that they have usually
employed them as perfect substitutes. In that context, both were thought of as describingunambiguously codifiable and, thus, decodifiable- signals about states of the world,
which agents could use merely as inputs in their maximizing calculations. As a
consequence, this view of information and knowledge excluded from the analysis issues
like learning processes, ambiguous signals, socially shared knowledge, uncodifiable (i.e.tacit) knowledge, and so on.
Although such issues have kept coming back to the forefront of investigations for
the whole of the past century, they have never successfully replaced the dominant view ofknowledge and information that I described above. However, my central argument in this
paper is that there seems to be a clear shift in economic theory in the last couple ofdecades, away from this dominant view and towards more sophisticated accounts of
knowledge. My aim here is not, of course, to survey the whole of 20th
century economics,
nor even to highlight all the important ideas on knowledge and information that various
economists put forward at various instances. My account will be highly selective, as itwill focus on those debates and theoretical strands that, to the authors view, have
exercised the greatest influence on the discipline, and will attempt to bring out the views
on knowledge and information implicit in each one of them. I argue that economists have
increasingly sought more sophisticated accounts of knowledge and information the more
they were involved in a sort of economics that tended to distance itself from price theory.Thus, we find consistently more elaborate contributions to the problem of knowledge astheory moved away from prices towards contracts and then towards institutions and
organizations.
The above imply that I adopt here a priori a different view of knowledge from the
dominant view. I will employ this view as a yardstick, to assess the extent that various
schools of thought measure up to it. For this purpose, I adopt here the taxonomy
introduced by Hodgson (1999). First of all, Hodgson (p. 46) distinguishes between sensedata, the various signals received by human senses, from information, which he describes
as data to which some meaning has been attributed. Importantly, different people may
attribute different meanings to the same piece of information. Hodgson describesknowledge as the product of information use. The information the agent receives,
together with the meaning she attributes to it, allow her to obtain knowledge. Therefore,while information is an input for the acquisition of knowledge, the latter is a wider and
deeper concept, for it constitutes both the prior endowment needed to make sense of any
information, as well as the final outcome of a process of learning. Finally, bothknowledge itself as well as the process leading to it can be either codifiable or tacit.
1
1 Hodgson, of course, follows here Polanyis (1967, p. 4) famous remark that we can know more than wecan tell.
7/30/2019 Xx Siecle Economics
3/21
3
Codifiable knowledge can be easily transferred at a cost- whereas tacit knowledge is not
easily alienable from its, individual or collective, bearer.
The paper is organized as follows. Section 3 briefly describes the view of
knowledge of standard neoclassical microeconomics, to which we referred above. Section
3 discusses the socialist calculation debate of the 1930s. Although this is almost an
entirely forgotten instance in the history of the discipline, its importance for the analysishere is that it brought out some crucial issues regarding the nature and the management of
knowledge, which were formally addressed by economics only decades later. Section 4
discusses a debate that never was, between Stigler and Arrow on the nature of
knowledge and information. Arguably, the opposing views of the two theorists introducea significant distinction between information as a commodity and knowledge as
something different from information. Section 5 analyzes briefly the views on knowledge
of New Classical Economics and of Endogenous Growth Theories. It is argued that both
strands revert to the dominant view of knowledge, as they treat it as an entirely codifiableentity. Section 6 discusses the importance of the concept of asymmetric information and
shows that the solution to the problems that it raises led theorists to the notion ofcontracts. Section 7 discusses various strands of New Institutional Economics and shows
how the various contractual theories of economic organization explain the emergence of
the capitalist firm as an institution. Section 8 discusses briefly the Resource-Based or
Capabilities of economic organization and argues that it is here that we find the mostsophisticated on the basis of the yardstick proposed above- treatment of knowledge in
economic theory. Finally, Section 9 sums up the papers conclusion.
2 The Problem of Knowledge in Standard Neoclassical Microeconomics
For whole generations of students of economics, the diagram of supply and demand2
hasbeen the entry point to the discipline. This is true even today, as it is evident from a
cursory inspection of most contemporary microeconomics textbooks. It is in that sense
that I refer to standard neoclassical microeconomics in the title of this Section. What I
wish to show here is that, in the context of that conceptualization of the workings of atypical market, knowledge is not indeed, it cannot- be approached as a problem for
theory but that it is reduced instead to an assumption that is essential to the functioning of
the model.
Every student of introductory microeconomics is expected to grasp the intuitivevalidity of the demand and supply schedules. Both schedules are supposed merely to
reflect the intentions of agents,3
about what quantities of goods to trade at each specific
price. Obviously, if there is an issue of knowledge here, it merely refers to the knowledgeby the agent of her own preferences, which she is supposed to possess given the
2 Usually referred to as the Marshallian scissors, after Alfred Marshall (1890) who introduced it more
than a hundred years ago.3 And, of course, the intentions are themselves reflections of the agents preferences.
7/30/2019 Xx Siecle Economics
4/21
4
rationality assumption. Of course, things are more complex in the case of the supply
schedule, which does not stem from the preferences of consumers but from the profit
maximizing decisions of firms.4
Still, the fact that both schedules reflect simply theintentions of agents may allow us to treat them as inward-looking concepts that simply
reflect what, from the agents point of view, is already given. In that sense, both the
demand and the supply schedules are entirely static concepts, as they reflect the agents
intentions on the basis of what they already know, rather than a plan for action towards anuncertain future.
However, the problem of knowledge emerges the moment we stop treating the
two schedules as isolated phenomena and we use them jointly in order to explain theoutcome of market interaction. From the technical point of view, we describe the point of
intersection of the two schedules as the point of market equilibrium, as any disturbance of
that state will automatically and instantaneously- tend to reinstate equilibrium. In
addition, and importantly for our point here, the intersection of the two curves determinesa Pareto-optimal state, so no agent has an incentive to change his/her behavior from that
which is consistent with that state. Again of course, the static nature of the model iscrystal clear.
The consistency of the model requires, therefore, the existence of some
mechanism that ensures the constant existence or, more generally, the constant tendencytowards the state of equilibrium. Not surprisingly, no such mechanism has ever been
proposed. The protagonists of the marginal revolution tried to overcome the problem
by proposing either a fictitious auctioneer (Walras 1874) or the idea that no trading
takes place before the establishment of equilibrium (Edgeworth ). However, even more
modern research has not been able to explain the process towards equilibrium, as it wasalready evident in Arrows (1959) attempt to explain the convergence of prices.
Given these difficulties, standard neoclassical microeconomics chose to focus
instead on the end state of that process, i.e. the state of equilibrium. To be conceptualizedas a truly end state, theorists had to assume that any tendency to disrupt it has already
ceased to operate. The assumption that provided the basis for such an approach was
perfect competition: the hypothesis that all competitive moves by agents have already
been made so that no further gains from trade are possible. A full list of the conditions forperfect competition can be already found in Frank Knight (1921).
5However, it was only
late in the interwar years6
and especially in the post 2nd
WW period that the concept of
perfect competition came to be fully accepted by the profession as an analytical tool,mostly as a result of the abandonment of Marshallian partial equilibrium analysis in favor
4 Arguably, the problem of knowledge begins to appear here, as the assumption of profit maximization is
based on the idea that the firm knows the prices of inputs, the available technologies, the prices of outputs,
and so on.5 However, it is interesting to note that Knight proposed that long list of conditions because he wanted to
focus on what happens in the event that some of them importantly, the perfect knowledge assumption-cease to hold. By contrast, later theorists, e.g. Stigler (1957), took Knights list as the final word on the
issue, i.e. as describing a real state towards the market always converges. For the history of the concept of
competition see McNulty (1967 and 1968).6 Hicks (1939) is a major landmark in that development.
7/30/2019 Xx Siecle Economics
5/21
5
of general equilibrium. The former, by offering the possibility to approach market
phenomena in terms of consecutive stages, allowed the partial investigation of the effects
brought about by the cancellation of various conditions of perfect competition. This wasthe case in the interwar years with the work of Robinson (1933/1979) and Chamberlin
(1933/1962), on imperfect and monopolistic competition respectively. By contrast, in
the state of general equilibrium all effects of competition are assumed to have worked
themselves out.
The concept of perfect competition has not, of course, escaped criticism from
various strands of heterodox economists. The suppression of time, implied by the static
nature of the analysis is one such criticism. The fact that the concept of perfectcompetition was expressly introduced in order to avoid the problem of the process of
market phenomena means that, under perfect competition, markets have to be assumed to
work instantaneously towards the establishment of general equilibrium after an initial
disruption of the original state. Second, and more importantly for my purpose in thispaper, a further criticism is that the concept of perfect competition precludes any
possibility of standard neoclassical microeconomics to theorize issues likeentrepreneurship, innovation and technological change i.e. the major instances in which
the emergence of new knowledge impacts on the development process- as it inevitably
depicts agents as constantly having to choose among a given array of fully known
alternatives.
In fact, there is one condition of perfect competition that seems to capture fully
the implications of such criticisms: the condition ofperfect knowledge. For perfect
competition to hold, agents are assumed to know fully the technological alternatives, the
prices of inputs and outputs, the best way to go about to achieve their objectives, and soon. Moreover, agents are assumed to possess a stock of knowledge that has been piled uphistorically i.e. its existence lies beyond the scope of theory- and which is common to
all. It is assumed that all new information that the system produces is immediately and
costlessly- available to all market participants and that it is interpreted and evaluated inthe same way by all. Therefore, and in contrast to the title of this Section, for standard
neoclassical economics knowledge is not a problem at all; it is simply assumed away.
3 Knowledge and the socialist calculation debate of the 1930s
The problem of the neoclassical treatment of knowledge came forcefully to the fore in the1930s, in a debate on the feasibility of socialism. One side in the debate, which Karen
Vaughn (1980) has labeled as neoclassical socialists, argued that a system of
comprehensive central planning was perfectly feasible and that its operation could beunderstood on the basis of the neoclassical conception of general equilibrium. Their
opponents from the liberal side of the political divide argued that socialism would be an
inherently inefficient form of social organization that would inevitably lead to
7/30/2019 Xx Siecle Economics
6/21
6
deterioration of living standards and political totalitarianism.7
Interestingly, although my
analysis will show that the issue of knowledge was the central problem around which the
debate revolved, few of the protagonists realized it at the time.
The opening shots were fired by Ludwig von Mises (1920), who attacked the
conception of socialism put forth in the Marxian tradition. For the latter, the major feature
of that form of social organization would be the abolition of private property in the meansof production. But, according to Mises, the abolition of private property would inevitably
abolish all markets for means of production, so how would their prices be formed?
Inevitably, they would have to be set by a central authority (the Central Planning Board:
CPB), in which case nothing could ensure that they would reflect the true valuations ofagents, i.e. the true scarcities of means of production. For this reason, according to
Mises, these fictitious prices would disseminate misinformation to the managers of
socialist firms and, as a consequence, the latter would mis-calculate hence the term by
which the debate came to be known- the true costs of production. Thus socialism wouldbe an inherently inefficient system of social organization that could not be viable in the
long run.
The rebuttal of Mises criticism from the socialist side came from a number of
economists (Dickinson 1933 and 1939, Lerner 1934-5, 1937, Durbin 1936), who claimed
that the solution to the problems he raised could be established by mathematical means,thus prompting Hayek to label this approach the mathematical solution. These theorists
sought to find ways in which the profit-seeking motives springing from private property
could be replaced by the conscious management of the economy by the CPB. The core of
this approach was perhaps best expounded by Dickinson (1933, p. 239), who demanded
that all firms of the socialist system operate within glass walls, i.e. that they be obligedto publish all information relating to their operation. It is precisely the public nature ofthis knowledge that would allow the CPB to collect all the data required in order to build
a mathematical model of the economy as a system of simultaneous equations. By solving
this system, the CPB would establish the market clearing prices and quantities, thusensuring general equilibrium.
The protagonist in the next stage of the debate was Friedrich Hayek, who pointed
out that the knowledge requirements of the mathematical solution would make itimpracticable. If the CPB attempted to set up the system of equations, the first problem it
would have to deal with is, according to Hayek (1935, p. 211), the acquisition of
knowledge of consumers preferences. However, the fact that preferences change all thetime means that the CPB would have to be engaged in a continuous process not just of
solving the system but also of constantly respecifying its parameters. Furthermore, theaccurate specification of the systems equations requires that the CPB obtains knowledge
about the most effective use of every factor of production. Evidently, the centralization of
this information is again impracticable because of its sheer volume, thus leading to theinefficient use of social resources.
7 See Ioannides (2000) for a survey of the debate.
7/30/2019 Xx Siecle Economics
7/21
7
The neoclassical socialists took Hayeks point about the practical difficulty of the
centralization of knowledge very seriously, which is very well reflected in Oscar Langes
(1938) model of a competitive socialist system, which marked an important break withthe models based on the mathematical solution. In Langes model, socialist enterprises
are allowed to compete amongst themselves on two conditions. The first is that managers
must observe some simple operational rules: minimize average cost and determine
production level at the point that the marginal cost equals the price. The second conditionis that prices and production levels for each industry are set be the CPB. However, in
contrast to the mathematical solution, prices are not to be set mathematically but
empirically. In effect, Lange proposes that the CPB operates as a sort of Walrasian
auctioneer: lowering prices in the markets with excess supply and raising them in themarkets with excess demand. The ex post nature of the CPBs intervention avoids the
problems of the mathematical solution, which attempted to affect price setting ex ante.8
It was again Hayek, who fired the final shots in the debate. In his view, whatmakes the system of central planning impracticable is not only the fact that the
centralization of individually held knowledge is difficult but, more importantly, thatalarge part of this knowledge is uncentralizable because of its tacit character. Hayek
(1963, p. 61) claims that a large part of individually held knowledge issupra conscious
and, as such, it cannot be alienated from its physical bearer. To this he added a further
argument, which was based on an idea that was later to become prominent in economicsunder the name ofopportunism. By its very nature, Langes price-setting method cannot
adjust prices to changing conditions as speedily as when private entrepreneurs are
allowed to negotiate prices and strike deals independently of any central direction. Thus,
what Hayek (1940, p.191) terms the period of fixation is generally longer than the
corresponding period in a freely operating market system. However, the question is howmanagers can be assumed to behave in that period. The logic of Langes proposal is thatmanagers continue to trade at current prices even if they expect that the CPB will have to
change them in the future. But how can it be ensured that the managers will loyally obey
the CPB rather than attempt to take advantage of their own knowledge opportunistically?
Therefore, although the issue of knowledge was not initially on the agenda of the
debate, it kept coming to the forefront as the debate evolved. More importantly, however,
-and with hindsight- the two sides seem to have been involved in an exercise in whicheach of them kept refining their views on knowledge, with little cross reference.
Interestingly, Mises argument made no reference to the problem of knowledge, other
than reconfirming the neoclassical view of equilibrium prices as bearers of correctinformation. On the other hand, both the supporters of the mathematical solution as well
as Oscar Lange remained trapped in a view of knowledge exclusively as codifiedinformation, with the consequence being that their continued revisions of their socialist
models aimed merely at addressing the practical aspects of the problem. By contrast, and
again with hindsight, Hayek seems to have made the most significant contributions, interms of the concepts of tacitknowledge and opportunism. However, their significance
was only to be recognized decades later and, even then, not fully.
8 Interestingly, Lange (1967) continued to defend the practicability of his model even much later, claimingthat the electronic computer would make the calculations of the model much easier.
7/30/2019 Xx Siecle Economics
8/21
8
For the profession at large, the neoclassical socialists won the debate, as Bergson
(1948) declared. Mises and Hayek were barely mentioned from then on. Their role in thedebate was thought to be merely equivalent to that of a sparring partner: they raised
difficult questions, thus helping their opponents to refine their tools. Post war economics
took no notice of the issues of tacit knowledge and of opportunism that were raised by
Hayek at the closing of the debate. The latter was to be rediscovered decades later and tobegin a prominent career in economic theory. The former had to wait even longer and it
remains much more controversial even today.
4 On the commodity character of information: a latent debateThe issue of knowledge reemerges in economics in the early 1960s. In this Part I will
examine two important works of that period. Interestingly, although they put forth
opposing views on the role of knowledge in economics, there is surprisingly little cross-reference. But how come these novel treatments of knowledge appeared simultaneously?
What was the reason for the sudden shift of economists towards the study of knowledge,
after a long period of time in which it was not thought of as a problem of any analyticalinterest?
The 1950s have been a very important period in the history of the discipline, as itwas in that decade that the General Equilibrium (GE) paradigm became established as
the dominant paradigm in economics.9
In fact, as many historians of the discipline havepointed out,
10the 1950s is the period in which one can talk about the emergence of an
orthodoxy in economics, an unprecedented development compared to the pluralism ofprevious decades. However, despite the general acceptance of GE, economists were not
blind to its severe limitations, foremost among them being the assumption of the
existence of perfect contingent market for all future goods and services. Indeed, GE only
makes sense if agents are assumed to be able to trade not just current goods but also
future goods, each of them characterized by two attributes: the instance in time that theactual trade is to take place and the state of the world that will prevail at that instance. In
that context, therefore, the assumption of contingent markets implied that agents
preferences extended fully into the future, i.e. that each agent is in the position to fully
specify the future trades she will wish to carry out at every conceivable state of the world.
To non economists, the assumption of contingent markets may sound absurd. Toeconomists trained in neoclassical microeconomics, however, it merely meant an
extension of the perfect knowledge assumption that we discussed in Part 2 above, and
which they had learned to accept as an analytically legitimate conceptualization of howmarkets actually work. Of course, the fact that it was an extension of a familiar postulate
9 With Arrow and Debreu (1954) acquiring the status of the paradigms Bible. Of course, the foundations
had been established much earlier by Hicks (1939) and especially my the neoclassical synthesis of Hicks
(1937) and Samuelson (..), that cancelled the anti-optimization thrust of Keynesian economics.10 See, for example,
7/30/2019 Xx Siecle Economics
9/21
9
could not mean that it would be unquestioningly accepted. Rather, it was precisely the
fact that the assumption of contingent markets constituted an extension of the postulate of
perfect knowledge that prompted the most prominent of them to attempt to address theacquisition of knowledge as a legitimate problem of economic inquiry.
One early such attempt, which initiated a strand in economic theory known as
the economics of information, is George Stiglers (1961) paper by the same title.Stiglers conceptual innovation is to treat knowledge as a commodity. In the context of
GE theory, the ingenuity of this innovation is, of course, obvious. By treating knowledge
as a commodity, Stigler manages two things: first, he fleshes out, as it were, the models
contingent markets by giving them an object to trade in today, rather than at some distantdate in the future. Secondly, by describing knowledge as a marketable commodity he can
proceed to analyze the market for knowledge with the well established tools of
neoclassical microeconomics. There is a further implication, which is of great interest for
what I am going to discuss later in this paper. The fact that knowledge is viewed as acommodity allows Stigler to treat it as something external to the agent, which the latter
neither produces nor can in any way affect through her action. There is no betterrealization that Stigler indeed has this view from the fact that, although he begins the
paper by using the terms knowledge and information interchangeably, he goes on to
focus on information.
Stigler maintains that the framework he is proposing can, in principle, cover all
kinds of information. However, he chooses to focus on just one such type: information on
prices. He begins his analysis by postulating a pre-equilibrium situation, in which no
uniform price for a specific commodity exists. This means that agents are ignorant of the
correct price for that commodity. In turn, this implies the existence of price dispersion,which is only another way of describing the ignorance that agents face. Indeed, Stigler(1961, p. 214) maintains that price dispersion is a manifestation and, indeed, it is the
measure- of ignorance in the market. Faced with this dispersion, rational agents have an
incentive to try to get informed about the prices offered, through a process that Stiglerdescribes as search. By searching for prices, the agent achieves two things: first, she
increases the stock of knowledge in her possession and, second, she reduces the price
dispersion in the market.
However, search itself has a cost, therefore the agent will engage in it only to the
extent that the expected benefits i.e. the possibility of discovering a lower price for the
commodity- are greater than the search-cost itself. On these grounds, Stigler maintainsthat the market for information is analyzable by the same tools that theory has
traditionally employed in the analysis of markets for goods and services: agents willengage in search up to the point that marginal benefits equal marginal costs. The
reaffirmation of the marginal equilibrium condition has a further important implication:
agents will never be fully or perfectly informed and, thus, some dispersion will alwaysremain. However, the remaining dispersion will be itself optimal, in the sense that agents
have rationally decided not to expend any resources in order to reduce it further.
7/30/2019 Xx Siecle Economics
10/21
10
Only a year after the publication of Stiglers paper, Kenneth Arrows Economic
Welfare and the Allocation of Resources for Invention (1962a) appeared. The central
issue, of course, is again knowledge. Unlike Stigler, however, who focused on theinformation conveyed by prices, Arrow concentrates on the creation of new knowledge,
through the process that the term invention usually describes. Importantly, therefore,
and in contrast to Stigler, Arrow focuses in the production rather the reception of
knowledge. On the other hand, importantly again, Arrow still refers to this newknowledge as information. It must be noted that, implicitly,
11Arrow deals only with
codifiable knowledge.
Arrow makes no reference to Stigler (1961). However, one cannot but feel that hispaper constitutes a direct critique of Stigler, as it aims to refute the latters central idea:
that information can be treated just like any other good or service. Against this view,
Arrow puts forth three interrelated arguments. The first is that the benefits that ensue
from a piece of information i.e. codified knowledge- are independent of the costs that itsreproduction entails. Thus, it is virtually impossible for the seller of a piece of
information to recover the full costs of its production, since the buyer can reproduce anddistribute it at low cost. On these grounds, the value of a piece of information will always
be much higher than the payment that its original seller will be able to command. The
seller will, therefore, always be in a disadvantage that can only partially be protected
through legal means e.g. patents, license contracts, etc.
Arrows second argument is that information is a peculiar kind of commodity also
in another sense: even after its sale, it never leaves the property of its original owner.
Obviously, it is now the buyer of the information that is placed at a disadvantage, for she
pays for a commodity without being able to exercise a full property right over it, since theoriginal owner remains an owner of the commodity. Again, one can think of a wholearray of legal means through which the buyer may attempt to diminish the original
owners right of use of the specific information. However, given the peculiar nature of
information as a commodity, such protection can never be absolute. On the other hand,the fact remains that the incentives to the buyer to buy the information are significantly
distorted and, as a consequence, she will wish to pay less than the actual value of the
information to her.
Arrows third argument focused again on the peculiar nature of information as a
commodity. For a normal commodity, price theory teaches that the buyer will be willing
to pay a price equal to the marginal utility she expects to enjoy from the consumption ofthe commodity. Implicit in this is the assumption that the buyer can fully ascertain her
marginal utility, for only if that is the case will she be in a position to know what shewishes to pay for it. However, the paradox of information as a commodity is that one
cannot know the utility derivable from it unless she already possesses it; but in that case
why buy it in the first place?12
On the basis of these arguments, Arrow concludes that the
11 As we will see below, Arrow (1962b) was one of the pioneers in introducing learning and, with it, the
tacit dimension of knowledge in economic theory.12 One can find a similar argument in the most prominent member of the modern Austrian school ofeconomics (Kirzner 1973).
7/30/2019 Xx Siecle Economics
11/21
11
costs of producing new knowledge in an uncertain context -through research- are subject
to high inappropriabilities, the consequence being that outlays in R&D will tend to be
lower than the social optimum. In fact, Arrows arguments on knowledge have furnishedan influential argument for the public funding of research, in most of the relevant
literature.13
Stigler and Arrow, therefore, offer two opposing views on the essence and the roleof information. The former insists on the commodity-like character of information, the
consequence being that the provision of information is thus described as an entirely
market outcome. In contrast, by denying the commodity character of information, Arrow
sees its provision as subject to market failure. However, the differences between the twocontributions are much deeper than that. For Stigler, the commodity character of
information implies a view of the latter as an unambiguous signal, which says the same
thing to all its recipients. In this context, the only difference between agents arises from
differences in their expected utilities from the same bit of information, which accountsfor the different costs they are willing to incur in order to obtain it. By contrast, we find
in Arrow one of the first instances in which the identity of the parties exchanginginformation matters. This is already obvious in the first two of his arguments, as it is clear
that it is important both to the seller and to the buyer of information to know what the
other party will do with it after the completion of the exchange. Therefore, the knowledge
of what the other party will do with a piece of information becomes a vital factor in theanalysis. It is also obvious in Arrows third argument, which puts into question the
rationality of any exchange of information. Clearly, such an exchange can only be
rational if we assume that the buyer of a piece of information already possesses a prior
knowledge, which allows her to assess the value of this information even prior to actually
obtaining it. Therefore, it is easily understood why Arrows contribution is generallyrecognized as having introduced a much richer conception of knowledge in economics.
5 The consolidation of the distinction between information and knowledge:
Rational Expectations and Endogenous Growth Theories
The 1970s has been the decade of the revolt against Keynesianism both at the theoretical
as well as at the policy level. In the former, the revolt has taken the form of the demand to
produce solid microfoundations for macroeconomic analysis. Milton Friedman hadpioneered the critique of Keynesianism in the previous two decades;
14however, it was in
the 1970s that the most devastating assault on Keynesian theory gained wide appeal
among economists under the name of New Classical Economics (NCE). Importantly for
our task here, the issue of the acquisition of knowledge lay at the foundation of that
paradigm.
13 An earlier statement to this effect can be found in Nelson (1959).14 His permanent income hypothesis (Friedman ..) and, especially, his attack (1968) on the Phillipscurve are evidence of this.
7/30/2019 Xx Siecle Economics
12/21
12
On the ground of sound neoclassical microeconomic principles, NCE assumes
that agents respond to changes to relative prices. But in order to be able to make sense of
the chaos that the constant changes in relative prices signify, the agent is also assumed tocontinuously compare these changes to changes in the general price level. In other words,
relative prices and the general price level constitute the major sources of information, on
the basis of which the agent decides her action.15
However, the actions of agents are
always directed at the future, so their actions are determined by the expectations theyform on the basis of the knowledge they obtain from price signals. But how are these
expectations formed? It is here that NCE introduces its greatest theoretical innovation:
the assumption ofrational expectations.16
In its stronger version, the assumption of rational expectations postulates that the
subjective probability an agent assigns to a future contingency matches exactly the
objective probability distribution.17
Thus the agent has already discounted accurately any
future eventuality by having already planned her action in a way that cancels anypossibility that the future may have any real consequences for her position. In the weaker
version of the rational expectations assumption, agents are thought of as responding toexternal events on the basis of a theory, which, given their extensive experience with
the functioning of the economy in the past, lead them to predict correctly the outcome of
any change in the systems parameters.18
Obviously, in both its versions the rational
expectations assumption amounts to a complete reaffirmation of the perfect knowledgepostulate of neoclassical micro-theory. The knowledge possessed by agents is thus
described as perfect in a dual sense: first, agents are supposed to have access and, thus, to
know perfectly the information contained in relative prices and in a set of macroeconomic
indices (inflation, unemployment, interest rates, etc.); second, their cognitive abilities are
such that allow them to process this information in the same way as all other agents and,thus, come to identical conclusions regarding the future prospects of the economy.
The high point of NCE was the period that followed the stagflation of the mid
1970s, a period in which the demand for macroeconomic stabilization was the order ofthe day. The 1980s and especially the 1990s, however, brought forth a new demand: the
explanation of long-term growth. The fact that this was a period of rapid and radical
technological development mainly in the area of Information and Telecommunication
Technologies- meant that this development had to be taken seriously into consideration asone of the determinants of growth.
19The new stream of research did exactly this, by
treating the production of new technological knowledge as an endogenous creation of the
system and has thus come to be known as Endogenous Growth Theories.
In order to endogenize technological change, various models proposed in thecontext of this stream lift some of the key conditions for the existence of perfect
15 See Lucas (1977, pp. 232-3) and Kyun (1988, p. 4).16 First introduced by Muth (1961).17
See Lucas (1977, p. 233).18 See Lucas (1979, p. 213) for the distinction between the strong and the weak versions of rational
expectations.19 In contrast to the approach to growth that was dominant in the 1950s, when technology was considered to
be entirely exogenous -the famous Solows residual. See Solow (1957).
7/30/2019 Xx Siecle Economics
13/21
13
competition; the model proposed by Grossman and Helpman (1991), for example,
postulates monopolistic competition, while OECD (1992) assumes the existence of
increasing returns. Technological change itself is viewed as the result of investment inknowledge creation, either by the private sector firms R&D outlays and individual
investment in human capital- or by the public sector investment in education and
research infrastructure. But what is it exactly that this investment is expended on? Most
models assume that these outlays buy new knowledge, which is thought of new codifiedand entirely approbriable information (Lipsey, 1999, p. 13). However, a small part of
endogenous growth models (Romer, 1990; Aghion and Howitt, 1992) stress the
inapprobriable character of knowledge and describe new technological knowledge as the
creation of a quasi-common (i.e. social) pool of usable information.
To some extent, therefore, the two streams of models reproduce the opposing
views on the character of information of Stigler and Arrow that we discussed in the
previous Part. This is certainly true on the issue of approabriability, as we have just seen.However, both streams treat knowledge as a stock, the quantity of which changes in
proportion to the investment that occurs. As important as the distinction betweenapprobriable and inapprobriable knowledge may be, therefore, both streams conceive of
knowledge as entirely codifiable. Thus, for both streams, knowledge is perceived as a
good, the difference between them simply having to do with its exact character: whether
private or public.
It is generally recognized that NCE and Endogenous Growth Theories constitute
the two most important and influential- developments in macroeconomics of the last
three decades. In terms of the problem of knowledge that interests us here, however, they
seem merely to reproduce the dominant view of knowledge and information as wedescribed it in the Introduction. In NCE we find a strong albeit implicit- recognition ofthe importance of prior knowledge, for the agent is thought of, in effect, as already
possessing the true model of the economy, so that they can instantly calculate the effects
of any exogenous disturbance. Therefore, agents can be described as behaving in acontext of perfect knowledge, for the simple reason that the theory has a priori assumed
that they possess this all-powerful prior knowledge. By contrast, Endogenous Growth
Theories treat knowledge as something augmentable through conscious and, of course,
rational- investment decisions. So what the models endogenize is, in effect, the creationof new knowledge. Again, however, this new knowledge is thought of as a stock of
codified information, which is accessible by all market participants at relatively low cost
and which, most importantly, has the same meaning for all.
6 Information asymmetries and the concept of contract: the return of
opportunism
The 1970s witnessed another major conceptual innovation, this time in micro-theory. TheArrow-Debreu model assumed that agents possess identical information about the current
7/30/2019 Xx Siecle Economics
14/21
14
and the future states of the economy. Uncertainty could and has been introduced in such
models but the assumption of identical information was retained in a sense, as agents
were described as being identically uncertain about future contingencies. The conceptualinnovation, to which I referred, constituted a clear break with this tradition, as it
introduced the concept of asymmetric information. Models were now constructed that
assumed that agents possess different sets of information, which were not mutually
revealed in the course of their market dealings. In fact, the asymmetry of informationprompted agents to act in ways that led to serious market imperfections.
20The fact that
agents know different things prompts each one of them to try to outguess the others, i.e.
to try to imagine what the others know and how they will act upon the knowledge they
possess. Thus, rather than being depicted as typical price-takers, agents are now thoughtof as behavingstrategically. In fact, the 1970s saw the wide adoption by economists of a
methodology that allowed the analysis of strategic behavior: game theory.
Clearly, the introduction of the concept of asymmetric information marks asignificant break with established micro-theory, as it implies an idea about information as
independent of prices. Information is now perceived as an integral part of the agentsinitial endowments, which she is in a position to manipulate strategically for personal
gain. However, unlike other kinds of endowments, non-disclosure constitutes the essence
of its effective exercise.21
Therefore, not only have agents no incentive to reveal this
personally held information, but they also have an incentive to act upon it in a way thatpromotes their own self interest against that of other agents, i.e. to act opportunistically.
Furthermore, asymmetric information does not only refer to the initial set of knowledge
possessed differently by various agents; it also refers to differential knowledge of agents
actions by the agents themselves and by other agents that merely observe the formers
actions.
On these grounds, the relevant literature usually distinguishes between two
categories of information asymmetries, although not all theorists agree on the criteria on
which the distinction must be based. I follow here Makowski and Ostroy (2001), whopropose a distinction between: a) asymmetries arising fromprivacy of information and b)
asymmetries arising from delivery problems. The first category refers to information that
is privately held by the parties and that is not revealed to others in the course of their
market dealings. In other words, this is a sort of information that could only be accessedhad other agents the possibility to look into the mind of its possessor. The second
category refers to information asymmetries between two agents that arise from the fact
that the first does not know how the second will act in a specific situation, i.e. what shewill actually deliver.
Both categories of asymmetric information pose formidable problems for market
efficiency. The latter depends on the coordinated action of agents; however, the existence
of asymmetric information implies that the prices that are formulated in markets
20 Akerlof (1970) is generally recognized as having initiated this stream of research.Other important
contributions to this stream of research are Holmstrm and Myerson (1983), Milgrom and Roberts (1982)
and, of course, more recent work by Joseph Stiglitz.21 Recall Arrows ideas on information as a commodity that we discussed in the previous Section.
7/30/2019 Xx Siecle Economics
15/21
15
characterized by asymmetric information cannot be considered as safe signals for
optimization. Therefore, the attainment of efficiency now requires the curbing of the
potential for opportunistic behavior that information asymmetries create. Of course, thesolution to the problem could not be sought in the formation of prices, so theorists had to
look for other ways in which rational agents could be thought of as solving the problem
of opportunism. Thus they focused on the concept of contracts: the instrument through
which agents choose to configure their dealings.
Contracts have traditionally been invoked in economics to refer to the legal means
and the legal form through which agents shape their exchanges of property rights over
goods and services. They could be explicit or implicit, the latter assumed to govern eventhe simplest transactions. Contracts were mainly thought as ensuring that the parties will
behave in the agreed manner thus, implicitly, the underlying idea was that of commonly
held knowledge, on the basis of which the parties agreed on joint future actions, and on
the sanctions in case of breaching. However, the possibility of opportunistic behavior bythe contracting parties now led to a major reconsideration of the notion of contracts in a
double sense. First, contracts should now be viewed as specifically aiming to curbopportunism, in situations characterized by significant information asymmetries. Second,
to the extent that such asymmetries characterized typical situations among transacting
parties, contracts were now viewed as giving rise to the creation of stable institutions.
7 Opportunism, contracts and institutions
These new developments inspired a novel strand of theorizing that explicitly addressedthe phenomenon of economic institutions and especially economic organization, which is
usually described as New Institutional Economics or, more generally, as the contractual
theories of economic organization. Although this strand is made up of various streams ofapproaches to the study of organizations and institutions, there exist some common
unifying principles that allow us to treat it as a unique paradigm. First, the concept of
contract constitutes the ultimate unit of analysis, as it is assumed that it is the instrument
through which agents attempt to structure their relations in ways that curb opportunism.Second, the possibility of opportunistic behavior by the parties to the contract is believed
to spring, first and foremost, from information asymmetries. Third, the process of
contracting among agents tends to establish stable organizations or institutions, in thesense that whole classes of institutions emerge that aim at mitigating the problem of
information asymmetries in specific kinds of contexts. Understandably, the institution-organization that has attracted the highest interest of economists working in that
paradigm is the business firm and other forms of business organizations.
The fact that contractual theories of economic organization begin with the
assumption of the existence of various sorts of information asymmetries means that they
can be usefully categorized into two streams, each stemming from the two kinds ofasymmetries according to Makowski and Ostroy (2001) that we mentioned above: a)
7/30/2019 Xx Siecle Economics
16/21
16
asymmetries arising from privacy and b) asymmetries arising from delivery problems.
The first stream includes transaction cost economics and the theory of incomplete
contracts, while the second is comprised by agency theory and the view of the firm as anexus of contracts. We will now briefly review these approaches to economic
organization and discuss the specific way in which they treat the problem of knowledge.
For transaction cost economics (Coase, 1937; Williamson, 1975 and 1985), afundamental transformation in contractual relations occurs as soon as two parties agree
on a contract that will guide their cooperation for a period of time in the future. Even if
the pre-contract situation were one of perfect competition, the post-contract situation is
transformed into one of bilateral monopoly. Thus, each party faces a potential hazardfrom the possibility that the other party will act opportunistically post-contractually.
22
Obviously, this possibility springs from the private knowledge of special importance
here is private knowledge obtained during contract execution- through which each party
will try to take advantage of any post-contractual contingency not explicitly addressed inthe contract.
23The possibility of opportunism is, of course, a major source of transaction
costs, as each party will have to take it into consideration when contemplating to enter along-term contractual relation. To avoid this cost, agents opt for a special kind of contract
that establishes a hierarchical relation between them, thereby replacing a market
transaction with a relation of fiat. Thus we have the emergence of the classical capitalist
firm as an institution.
The theory of incomplete contracts (Grossman and Hart, 1986; Hart, 1995)
focuses on ownership of specific resources rather on transactions, which is defined by
Hart (1995) as the residual control rights over assets .. the law. The importance of
ownership stems from the fact that any long-term contract will be inevitably incomplete,in the sense that it will never be possible to foresee all future contingencies. Thisincompleteness makes it necessary that the original contract setting up a firm assigns
ownership rights to the party that is most likely to make the necessary investment to
enhance the profitability of the organization as a whole. Obviously, this will be the partythat has the private knowledge about how to best pursue the joint effort. Therefore, both
for transaction costs economics as well as for the incomplete contracts theory of the firm,
it is privately-held knowledge that explains the emergence of the firm as an organization
and the organizational forms that it takes in various contexts.
As already mentioned, the second stream of contractual theories put delivery
problems at center stage. Agency theory (Ross, 1973; Jensen and Meckling, 1976) is onesuch approach that conceptualizes the principal-agent relation. In that context, a principal
is supposed to delegate a task to an agent to carry out on the formers behalf. Theprincipal is assumed to be unable to tightly monitor the agents work, so she has to judge
the agents work by assessing the latters performance. However, this assessment is not
unproblematic, as it is difficult for the principal to judge the extent to which the agentsperformance reflects her actual conduct or is a matter of sheer luck. On these grounds,
22 Thus Williamson (1985, p. 175) defines opportunism as self interest seeking with guile.23 Williamson assumes bounded rationality, which explains why it is not possible to address every possiblecontingency ex ante.
7/30/2019 Xx Siecle Economics
17/21
17
agency theory purports to construct models that devise efficient incentives structures,
thus making the agent behave in the principals best interest. Again, it is obvious that the
principal-agent problem arises from the information asymmetries that characterize theknowledge sets of each of the two parties, which translates into the possibility of the
agents opportunistic behavior. However, in this case, what the principal wants to know is
not the information the agent holds initially but, rather, the actual performance that the
agent will ultimately deliver.
The nexus of contracts view of the firm (Alchian and Demsetz, 1972) falls also
into the category of delivery problems arising from asymmetric information. The problem
here arises from the fact that, in the context of team production, where metering eachagents performance is costly, every agent has an incentive to shirk, i.e. to free-ride on the
efforts of other team members. Again, what is at issue is the actual performance that each
team member will ultimately deliver, regardless of the initial knowledge with which she
enters the employment contract. According to Alchian and Demsetz, the solution to thispeculiar kind of information asymmetry is the contractual arrangement of the monitoring
function, whereby one agent is assigned the responsibility (and, thus, the cost) ofmonitoring the performance of all others, in return for being recognized as the sole
residual claimant of team production. Again, we have here an account of the emergence
of the classic capitalist firm.
The above account clearly shows that contractual theories of economic
organization introduce into the analysis some of the most advanced notions of knowledge
and information that modern neoclassical economics has to offer. On these grounds, they
explicitly introduce one attribute of knowledge that had not appeared in economics since
the calculation debate: that it is liable to be treated opportunistically. We argue that thishas become possible because these theories have focused on contracts as the units ofanalysis, rather on price formation in markets. By focusing on interpersonal contracts,
they inevitably had to transcend the view of prices as the sole bearers of information and
to look deeper into the incentives that shape the intentions of agents to enter intocontracts.
8 From contracts to capabilities: the centrality of knowledge
The proponents of contractual theories of economic organization and of New
Institutional Economics more generally-24
pride themselves for the fact that their
paradigm attempts to study economic organization with the well established tools of
neoclassical microeconomics. This may add rigor to the analysis but, at the same time, it
imposes on it the strict static outlook of that paradigm. Foss (1998, pp. 182-3) offers aconcise list with three limitations of contractual analyses: a) an implicit assumption that
alternatives are given, thus depicting agents as having to choose among a very clearly
24 See how this paradigm is described by three of its main protagonists: Coase (1984), North (1986),Williamson (1985).
7/30/2019 Xx Siecle Economics
18/21
18
defined set of contractual alternatives; b) a suppression of process, which implies that the
optimal solution to the contract-design problem continues to be optimal throughout
contract execution; c) a set of strong knowledge assumptions, thus leaving no room fortheory to conceptualize the discovery of new knowledge by agents. Importantly, all three
drawbacks that Foss describes revolve around the issue of knowledge. The first implies a
view of knowledge as information that is already available to agents. The second
suppresses time as an essential aspect of the analysis, implying that agents do not learnanything during the execution of the contract. Finally, the third blocks any understanding
of the discovery of new knowledge by agents.
And yet, contractual theories of economic organization brought along two novelideas, thus opening a research path that attempts to break loose of the constraints that
these drawbacks represent. The first idea is that it is possible to transcend the traditional
view of economic organization, and the business firm more specifically, on the basis of
the neoclassical concept of the production function. Rather than viewing the firm as ablack box, a mere transformer of inputs into outputs, it was now recognized that
economic theory could open up the black box as it were, and could seek to explain therelations between its constituent elements. The second idea is that, by focusing on the
concept of interpersonal contracts rather than merely on market-mediated relations, it was
now possible to extend the analysis to the study of a wide range of hybrid forms of
economic organization like firm alliances, networks, and clusters. This line of researchallowed the abandonment of the view of competition as the sole determinant of inter-firm
relations, along the lines that Richardson (1972) had suggested.
This new line of research has developed in many directions, has focused on a
wide variety of organizational aspects, and has thus come to be described under variouslabels: Resource-Based View of the firm, Evolutionary Theory, Knowledge-BasedTheory of the Firm, to name only a few. However, in spite of the diversity of these
approaches, what arguably unifies this stream of research into one distinct paradigm is
the centrality of the concept ofknowledge, which, as we will argue below, acquires thefullest recognition of its attributes in economics.
A full reference to this already vast literature is not possible here. However, the
distinctiveness of the treatment of knowledge in the context of this paradigm werealready evident in the two contributions that are generally recognized as having laid its
foundations: Penrose (1959) and Nelson and Winter (1982). Penrose (1959) is generally
recognized as the initiator of the Research-Based View of the Firm, as she was the firsteconomist to describe the firm as a bundle of resources. But her greater contribution was
the distinction between the resources available to the firm and the services that the firmcan draw from them. In her view, there is not a one-to-one relation between the two
concepts, which means that the firm can obtain different services from different
combinations of its resources. But in this context, what is inside the firm becomes anobject of discovery by the firms management, as the services obtainable from different
combinations of resources are not immediately available but something for the
management to discover. Clearly, Penrose provides a forceful argument for the
7/30/2019 Xx Siecle Economics
19/21
19
importance of learning by doing and of the sources of innovation in the modern firm. In a
sense, entrepreneurship comes back to the forefront of economic theory after Schumpeter.
Nelson and Winter (1982) are generally recognized as the initiators of modern
evolutionary economics. Their greatest contribution is the introduction of the concept of
routines. Routines are products of human interaction within the firm, however, they are
mostly spontaneous growth of interactions, in the sense that they have not been designedby any single or collective will. As such, routines constitute a type of knowledge that is
characterized by two major attributes: it has a large part that it is tacit, while at the same
time it is social, in the sense that the knowledge represented by routines is not held by
individual agents but it is shared by all the human constituents of the firm.
References
Akerlof, G. (1970), The Market for Lemons: Quality, Uncertainty and Market Mechanism QuarterlyJournal of Economics, August, 488-500.
Alchian, A.A. and H. Demsetz, (1972), Production, Information Costs, and Economic Organisation,
American Economic Review, 62, 777-95.
Arrow, K.J. and Debreu, G. (1954), Existence of an Equilibrium for a Competitive Economy,Econometrica, 22, 265-90.
Arrow, K.J. (1959), Toward a Theory of Price Adjustment, in Abramovitz, M. et al. (eds), The Allocation
of Economic Resources, Stanford: Stanford University Press.
Arrow, K.J. (1962a), Economic Welfare and the Allocation of Resources to Invention, in Nelson, R.R.(ed.), (1962), The Rate and Direction of Inventive Activity: Economic and Social Factors,
Princeton N.J.: Princeton University Press.
Arrow, K.J. (1962b), The Economic Implications of Learning by Doing,Review of Economic Studies, 29,155-73.
Bergson, A. (1948), Socialism, in Ellis, H.S. (ed.), (1949),A Survey of Contemporary Economics, New
York: Blakiston.
Chamberlin, E.H. (1933/1962), The Theory of Monopolistic Competition, Cambridge: Harvard UniversityPress.
Coase, R.H. (1937), The Nature of the Firm,Economica, 4, 386-405.Coase, R.H. (1984), The New Institutional Economics,Journal of Institutional and Theoretical
Economics, 140, 229-231.
Dickinson, H.D. (1933), Price Formation in a Socialist Community,Economic Journal, 43, 237-50.Dickinson, H.D. (1939),Economics of Socialism, London: Oxford University Press.
Durbin, E.F.M. (1936), Economic Calculus in a Planned Economy,Economic Journal, 46, 676-90.
Edgeworth,
Foss, N.J. (1998), Austrian Insights and the Theory of the Firm,Advances in Austrian Economics, 4, 175-
98.Grossman, S.J. and O.D. Hart, (1986), The Costs and Benefits of Ownership,Journal of Political
Economy, XCIV: 671-719.
Hart, O. (1995),Firms, Contracts and Financial Structure, Oxford: Clarendon Press.
7/30/2019 Xx Siecle Economics
20/21
20
Hayek, F.A. (ed.), (1935/1975), Collectivist Economic Planning, London: G. Routledge & Sons.
Hayek, F.A. (1935), The Present State of the Debate, in Hayek (1935/1975).Hayek, F.A. (1940), Socialist Calculation III, in Hayek (1949).
Hayek, F.A. (1949),Individualism and Economic Order, London: Routledge & Kegan Paul.Hayek, F.A. (1963), Rules, Perception and Inteligibility, in Hayek (1967).Hayek, F.A. (1967), Studies in Philosophy, Politics and Economics, London: Routledge & Kegan Paul.
Hicks, J. (1937), Mr. Keynes and the Classics: A Suggested Interpretation,Econometrica, 5, 147-59.
Hicks, J. (1939), Value and Capital, Oxford: Oxford University Press.
Hodgson, G.M. (1999),Economics and Utopia: Why the Learning Economy is not the End of History,London: Routledge.
Holmstrm, B. and Myerson, R. (1983), Efficient and Durable Decision Rules with Incomplete
Information,Econometrica, 51, 1799-1820.
Jensen, M.C. and W.H. Meckling, (1976), Theory of the Firm: Managerial Behavior, Agency Costs, andOwnership Structure,Journal of Financial Economics, 305-360.
Ioannides, S. (2000), Austrian Economics, Socialism and Impure Forms of Economic Organisation,
Review of Political Economy, 12, 45-71.
Kirzner, I.M. (1973), Competition and Entrepreneurship, Chicago: University of Chicago Press.Knight, F.H. (1921),Risk Uncertainty and Profit, Chicago: University of Chicago Press.
Kyun, K. (1988),Equilibrium Business Cycle Theory in Historical Perspective, Cambridge: CambridgeUniversity Press.
Lange, O. (1938), On the Economic Theory of Socialism, in Lippincott, B.E. (ed.), On the Economic
Theory of Socialism, New York: McGraw Hill.Lange, O. (1967), The Computer and the Market, in Feinstein, C. (ed.), (1967), Capitalism, Socialism and
Economic Growth, Cambridge: Cambridge University Press.
Lerner, A.P. (1935-5), Economic Theory and Socialist Economy,Review of Economic Studies, 2, 51-61.
Lerner, A.P. (1937), Statics and Dynamics in Socialist Economics,Economic Journal, 47, 253-70.
Lucas, R.E. Jr. (1977), Understanding Business Cycles, in Lucas (1981).
Lucas, R.E. Jr. (1979), An Equilibrium Model of Business Cycle, in Lucas (1981).Lucas, R.E. Jr. (1981), Studies in Business Cycle Theory, Cambridge Mass: The MIT Press.
Makowski, L. and Ostroy, J.M. (2001), Perfect Competition and the Creativity of the Market,Journal of
Economic Literature, 34, 479-535.Marshall, A. (1990), The Principles of Economics, London: Macmillan.
McNulty, P.J. (1967), A Note on the History of Perfect Competition,Journal of Political Economy, 75,
395-9.
McNulty, P.J. (1968), Economic Theory and the Meaning of Competition, Quarterly Journal ofEconomics, 82, 639-56.
Milgrom, P. and Roberts, J. (1982), Limit Pricing and Entry under Incomplete Information: An
Equilibrium Analysis,Econometrica, 50, 443-59.
Mises, L. von, (1920), Economic Calculation in the Socialist Commonwealth, in Hayek (1935).Muth, J.F. (1961), Rational Expectations and the Theory of Price Movements,Econometrica, 29, 315-35.
Nelson, R.R. (1959), The Simple Economics of Basic Scientific Research,Journal of Political Economy,
67, 297-306.Nelson, R.R. and Winter, S.G. (1982),An Evolutionary Theory of Economic Change, Cambridge: Harvard
University Press.Nonaka ., and Takeushi, (1995), The knowledge-creating company, Oxford University Press: New York.
North, D.C. (1986), The New Institutional Economics, Journal of Institutional and Theoretical
Economics, 142, 230-237.Penrose, E.T. (1959), The Theory of the Growth of the Firm, Oxford: Oxford University Press.
Polanyi, M. (1967), The Tacit Dimension, London: Routledge and Kegan Paul.
Richardson, G.B. (1972), The Organisation of Industry,Economic Journal, 82, 883-96.
Robinson, J. (1933/1979), The Economics of Imperfect Competition, London: Macmillan.Ross, S.A. (1973), The Economic Theory of Agency: The Principals Problem,American Economic
Review, 63, 134-9.
Samuelson,
7/30/2019 Xx Siecle Economics
21/21
21
Solow, R.M. (1957), Thechnical Change and the Aggregate Production Function The Review of
Economics and Statistics, 39, 312-20.Stigler, G.J. (1957), Perfect Competition Historically Contemplated,Journal of Political Economy, 65, 1-
17.Stigler, G.J. (1961), The Economics of Information,Journal of Political Economy, 69, 213-25.Vaughn, K.I. (1980), Economic Calculation under Socialism,Economic Inquiry, 18, 535-54.
Walras, L. (1874/1954),Elements of Pure Economics, New York: Augustus Kelley.
Williamson, O.E. (1975),Markets and Hierarchies: Analysis and Anti-trust Implications, New York: Free
Press.Williamson, O.E. (1985), The Economic Institutions of Capitalism, New York: Free Press.
Williamson. O.E. (1985), Reflections on the New Institutional Economics,Journal of Institutional and
Theoretical Economics, 141, 187-195.